Movement path detection for anomalies and patterns from sensors in a home or other environment

Abstract

A method for detecting a movement path in a monitored environment generates a first movement path record by collecting first node information for each of a first series of nodes including at least one of a sensor ID or a device ID, and a trigger time. The method further determines a first movement path total duration time from a difference between a time of the ending node and a time of the starting node of the first series of nodes. The method further stores the first movement path record and subsequently compares a second movement path record with the first movement path record by comparing the first node information with second node information for each of a second series of nodes of the second movement path record, and the first movement path total duration time with a second movement path total duration time. The method determines a similarity value between the second movement path record and the first movement path record based on the comparison and detects a movement path based on the determined similarity value.

The invention claimed is: 
     
       1. A computer-implemented method for detecting a movement path in a monitored environment, the method comprising:
 generating a first movement path record by:
 collecting first node information for each of a first series of nodes, each first node information including at least one of a sensor ID or a device ID, and a trigger time, 
 determining a starting node and an ending node from the first series of nodes based on respective trigger times, and 
 determining a first movement path total duration time from a difference between a time of the ending node and a time of the starting node of the first series of nodes; 
 
 storing the first movement path record with a first movement path record weight value; 
 generating a second movement path record by:
 collecting second node information for each of a second series of nodes, each second node information including at least one of a sensor ID or a device ID, and a trigger time, 
 determining a starting node and an ending node from the second series of nodes based on respective trigger times, and 
 determining a second movement path total duration time from a difference between a time of the ending node and a time of the starting node of the second series of nodes; 
 
 comparing the second movement path record with the first movement path record by comparing at least one of:
 the first node information for each of the first series of nodes and the second node information for each of the second series of nodes, and 
 the first movement path total duration time and the second movement path total duration time; 
 
 determining a similarity value based on comparing the second movement path record and the first movement path record by performing unsupervised machine learning by at least one of graph data mining and association analysis on the first and second movement path records, wherein the unsupervised machine learning is performed remotely and off-line from the monitored environment; and 
 detecting a movement path in the monitored environment based on the determined similarity value. 
 
     
     
       2. A computer-implemented method for detecting a movement path in a monitored environment, the method comprising:
 generating a first movement path record by:
 collecting first node information for each of a first series of nodes, each first node information including at least one of a sensor ID or a device ID, and a trigger time, 
 determining a starting node and an ending node from the first series of nodes based on respective trigger times, and 
 determining a first movement path total duration time from a difference between a time of the ending node and a time of the starting node of the first series of nodes; 
 
 storing the first movement path record with a first movement path record weight value; 
 generating a second movement path record by:
 collecting second node information for each of a second series of nodes, each second node information including at least one of a sensor ID or a device ID, and a trigger time, 
 determining a starting node and an ending node from the second series of nodes based on respective trigger times, and 
 determining a second movement path total duration time from a difference between a time of the ending node and a time of the starting node of the second series of nodes; 
 
 comparing the second movement path record with the first movement path record by comparing at least one of:
 the first node information for each of the first series of nodes and the second node information for each of the second series of nodes, and 
 the first movement path total duration time and the second movement path total duration time; 
 
 determining a similarity value based on comparing the second movement path record and the first movement path record; 
 detecting a movement path in the monitored environment based on the determined similarity value; 
 at a first network-connected sensor or device, sensing a user presence proximate a first media presentation device and determining a media content identifier of media content that may be played back on the first media presentation device; 
 at a second network-connected sensor or device, sensing the user presence at a first time and storing a queue position of the media content that may be played back on the first media presentation device corresponding to the first time; 
 determining an anticipated movement path from a plurality of stored movement path records based on a similarity of values between one of the stored movement path records and the sensed user presence at the second network-connected sensor or device; 
 determining a destination of the anticipated movement path and a second media presentation device proximate the destination; 
 requesting the media content identifier of the media content and queue the media content at the queue position for playback at the second media presentation device; and 
 at a third network-connected sensor or device proximate the destination of the anticipated movement path, sensing the user presence and subsequently initiating playback of the media content at the queue position on the second media presentation device. 
 
     
     
       3. A computer-implemented method for detecting a movement path in a monitored environment, the method comprising:
 generating a first movement path record by:
 collecting first node information for each of a first series of nodes, each first node information including at least one of a sensor ID or a device ID, and a trigger time, 
 determining a starting node and an ending node from the first series of nodes based on respective trigger times, and 
 determining a first movement path total duration time from a difference between a time of the ending node and a time of the starting node of the first series of nodes; 
 
 storing the first movement path record with a first movement path record weight value; 
 generating a second movement path record by:
 collecting second node information for each of a second series of nodes, each second node information including at least one of a sensor ID or a device ID, and a trigger time, 
 determining a starting node and an ending node from the second series of nodes based on respective trigger times, and 
 determining a second movement path total duration time from a difference between a time of the ending node and a time of the starting node of the second series of nodes; 
 
 comparing the second movement path record with the first movement path record by comparing at least one of:
 the first node information for each of the first series of nodes and the second node information for each of the second series of nodes, and 
 the first movement path total duration time and the second movement path total duration time; 
 
 determining a similarity value based on comparing the second movement path record and the first movement path record; 
 detecting and recording a movement path in the monitored environment based on the determined similarity value; 
 determining whether a sensed signal is received within a time period proximate a time value of the movement path record; 
 performing one of
 incrementing a weight value of the movement path record when the sensed signal is received within the time period, or 
 decrementing the weight value of the movement path record when the sensed signal is not received within the time period; and 
 
 sending a power off signal to at least one of a network-connected lighting device or a network-connected appliance based on the sensed signal not being received within the time period. 
 
     
     
       4. The computer-implemented method of  claim 3 , wherein the sensed signal is received from a network-connected sensor. 
     
     
       5. The computer-implemented method of  claim 3 , wherein the sensed signal is received from a network-connected device. 
     
     
       6. A computer-implemented method for detecting a movement path in a monitored environment, the method comprising:
 generating a first movement path record by:
 collecting first node information for each of a first series of nodes, each first node information including at least one of a sensor ID or a device ID, and a trigger time, 
 determining a starting node and an ending node from the first series of nodes based on respective trigger times, and 
 determining a first movement path total duration time from a difference between a time of the ending node and a time of the starting node of the first series of nodes; 
 
 storing the first movement path record with a first movement path record weight value; 
 generating a second movement path record by:
 collecting second node information for each of a second series of nodes, each second node information including at least one of a sensor ID or a device ID, and a trigger time, 
 determining a starting node and an ending node from the second series of nodes based on respective trigger times, and 
 determining a second movement path total duration time from a difference between a time of the ending node and a time of the starting node of the second series of nodes; 
 
 comparing the second movement path record with the first movement path record by comparing at least one of:
 the first node information for each of the first series of nodes and the second node information for each of the second series of nodes, and 
 the first movement path total duration time and the second movement path total duration time; 
 
 determining a similarity value based on comparing the second movement path record and the first movement path record; 
 detecting and storing a movement path in the monitored environment based on the determined similarity value; 
 determining a location-based condition from a received signal; 
 determining a presence of a user is not proximate a location originating the received signal; 
 determining the user is one of currently on or has completed the stored movement path; and 
 alerting the user of the location-based condition based on determining the user is one of currently on or has completed the movement path corresponding to the stored movement path. 
 
     
     
       7. The computer-implemented method of  claim 6 , wherein the received signal is from a network-connected sensor at the location originating the received signal. 
     
     
       8. The computer-implemented method of  claim 6 , wherein the received signal is from a network-connected device at the location originating the received signal. 
     
     
       9. A system for detecting a movement path in a monitored environment, the system comprising:
 a memory configured to store processor-readable instructions; and 
 one or more processors in communication with the memory, the one or more processors configured to execute the processor-readable instructions to:
 generate a first movement path record by:
 collecting first node information for each of a first series of nodes, each first node information including at least one of a sensor ID or a device ID, and a trigger time, 
 determining a starting node and an ending node from the first series of nodes based on respective trigger times, and 
 determining a first movement path total duration time from a difference between a time of the ending node and a time of the starting node of the first series of nodes; 
 
 cause the first movement path record to be stored with a first movement path record weight value; 
 generate a second movement path record by:
 collecting second node information for each of a second series of nodes, each second node information including at least one of a sensor ID or a device ID, and a trigger time, 
 determining a starting node and an ending node from the second series of nodes based on respective trigger times, and 
 determining a second movement path total duration time from a difference between a time of the ending node and a time of the starting node of the second series of nodes; 
 
 compare the second movement path record with the first movement path record by comparing at least one of:
 the first node information for each of the first series of nodes and the second node information for each of the second series of nodes, and 
 the first movement path total duration time and the second movement path total duration time; 
 
 determine a similarity value based on comparing the second movement path record and the first movement path record by performing unsupervised machine learning by at least one of graph data mining and association analysis on the first and second movement path records, wherein the unsupervised machine learning is performed remotely and off-line from the monitored environment; and 
 detect a movement path in the monitored environment based on the determined similarity value. 
 
 
     
     
       10. A system for detecting a movement path in a monitored environment, the system comprising:
 a memory configured to store processor-readable instructions; and 
 one or more processors in communication with the memory, the one or more processors configured to execute the processor-readable instructions to:
 generate a first movement path record by:
 collecting first node information for each of a first series of nodes, each first node information including at least one of a sensor ID or a device ID, and a trigger time, 
 determining a starting node and an ending node from the first series of nodes based on respective trigger times, and 
 determining a first movement path total duration time from a difference between a time of the ending node and a time of the starting node of the first series of nodes; 
 
 cause the first movement path record to be stored with a first movement path record weight value; 
 generate a second movement path record by:
 collecting second node information for each of a second series of nodes, each second node information including at least one of a sensor ID or a device ID, and a trigger time, 
 determining a starting node and an ending node from the second series of nodes based on respective trigger times, and 
 determining a second movement path total duration time from a difference between a time of the ending node and a time of the starting node of the second series of nodes; 
 
 compare the second movement path record with the first movement path record by comparing at least one of:
 the first node information for each of the first series of nodes and the second node information for each of the second series of nodes, and 
 the first movement path total duration time and the second movement path total duration time; 
 
 determine a similarity value based on comparing the second movement path record and the first movement path record; 
 detect a movement path in the monitored environment based on the determined similarity value; 
 
 a first network-connected sensor or device, configured to:
 sense a user presence proximate a first media presentation device and determining a media content identifier of media content that may be played back on the first media presentation device; 
 
 a second network-connected sensor or device, configured to:
 sense the user presence at a first time and storing a queue position of the media content that may be played back on the first media presentation device corresponding to the first time, wherein the one or more processors are further configured to execute the processor-readable instructions to:
 determine an anticipated movement path from a plurality of stored movement path records based on a similarity of values between one of the stored movement path records and the sensed user presence at the second network-connected sensor or device; 
 determine a destination of the anticipated movement path and a second media presentation device proximate the destination; and 
 request the media content identifier of the media content and queue the media content at the queue position for playback at the second media presentation device; and 
 
 
 a third network-connected sensor or device, configured to:
 sense the user presence and subsequently initiating playback of the media content at the queue position on the second media presentation device. 
 
 
     
     
       11. A system for detecting a movement path in a monitored environment, the system comprising:
 a memory configured to store processor-readable instructions; and 
 one or more processors in communication with the memory, the one or more processors configured to execute the processor-readable instructions to:
 generate a first movement path record by:
 collecting first node information for each of a first series of nodes, each first node information including at least one of a sensor ID or a device ID, and a trigger time, 
 determining a starting node and an ending node from the first series of nodes based on respective trigger times, and 
 determining a first movement path total duration time from a difference between a time of the ending node and a time of the starting node of the first series of nodes; 
 
 store the first movement path record with a first movement path record weight value; 
 generate a second movement path record by:
 collecting second node information for each of a second series of nodes, each second node information including at least one of a sensor ID or a device ID, and a trigger time, 
 determining a starting node and an ending node from the second series of nodes based on respective trigger times, and 
 determining a second movement path total duration time from a difference between a time of the ending node and a time of the starting node of the second series of nodes; 
 
 compare the second movement path record with the first movement path record by comparing at least one of:
 the first node information for each of the first series of nodes and the second node information for each of the second series of nodes, and 
 the first movement path total duration time and the second movement path total duration time; 
 
 determine a similarity value based on comparing the second movement path record and the first movement path record; 
 detect and recording a movement path in the monitored environment based on the determined similarity value; 
 determine whether a sensed signal is received within a time period proximate a time value of the movement path record; 
 perform one of:
 incrementing a weight value of the movement path record when the sensed signal is received within the time period, or 
 decrementing the weight value of the movement path record when the sensed signal is not received within the time period; and 
 
 cause a power off signal to be sent to at least one of a network-connected lighting device or a network-connected appliance based on the sensed signal not being received within the time period. 
 
 
     
     
       12. The system of  claim 11 , further comprising a network-connected sensor, wherein the sensed signal is received from the network-connected sensor. 
     
     
       13. The system of  claim 11 , further comprising a network-connected device, wherein the sensed signal is received from the network-connected device.